116

Assessment of Geosynthetics Interface Friction for Slope Barriers of Landfill

J.P.Gourc , R.Reyes-Ramirez & P.Villard Lirigm, University Grenoble 1, France gourc@ujf-grenoble.fr

ABSTRACT: Stability of Geosynthetic Lining Systems is, for a geotechnical standpoint, a complex matter. Three geomechanical questions were identified: sliding of the geosynthetic lining system on slope, pull-out strength of the geosynthetic anchorage at the top of the slope, rain erosion of the cap cover. Research programmes carried out in France on these topics, , are presented. Use of laboratory facilities (mainly different Inclined Planes) and large scale experimentations on actual slopes is especially emphasized. The observations derived from the tests and their detailed interpretation are really fruitful, as they highlight specific local interaction behaviour between soil and geosynthetics, which are not taken into account in design methods, more particularly wrinkles and real relative displacements of geosynthetics along the slope (Chapter 2), realistic value of interface friction angle (Chapter 3 ), pull-out strength of a L-shape anchorage (Chapter 4), identification of the mechanisms of control of erosion by geosynthetics (Chapter 5).

1 INTRODUCTION

Preservation of the lining barrier of domestic and industrial waste disposal is important firstly for environmental reasons. One of the modern strategies for waste disposal is the concept of bioreactor (Figure 1): the landfill is now considered as a center of energy recovering , which requires still more care specifically for the implementation of the cap cover ( Olivier et al, 2003), since: - biogas should be collected with the minimum loss through the barrier - leachate could be recirculated in order to accelerate the waste degradation without uncontrolled water

supply from the cap cover. The control of flow through the cap cover induced serious geotechnical problems, because: - cap liner are more and more composite structures, Geosynthetic Lining Systems (GLS), with interface

matters - slopes of cap cover are steep, in order to make more profitable the disposal site, with the maximum dumping

volume. - the cap cover is supported by a waste body which is often extremely compressible.

Figure 1. The bioreactor new concept for an updated domestic waste landfill (from Environmental French

Agency Ademe )

117

This Lecture is the opportunity to summarize several research works carried out at the Lirigm of the University of Grenoble in France, related to the different stability problems arising in landfill applications. Three questions were identified (Figure 2): - sliding of the geosynthetic lining system on slope (chapters 2 and 3) - pull-out strength of the geosynthetic anchorage (chapter 4) - rain erosion of the cap cover This research programme includes theoretical and experimental development, but in the framework of this Lecture, use of laboratory facilities (mainly Inclined Planes) and large scale experimentations is emphasized.

Figure 2. Main issues related to the stability of Geosynthetic Lining Systems on landfill slopes

2 STABILITY OF GEOSYNTHETIC LINING SYSTEMS (GLS) ON SLOPE

In the framework of large research programmes, sponsored by the French Environmental Agency ADEME and Industrial Companies, two large scale experimentations were carried out on actual landfill sites (Montreuil and Torcy in France). Main results are presented below.

The sealing systems used for the sloping sides of waste storage centres are made up of different geosynthetic and mineral components. The distribution of forces within each component is complex and results mainly from the deformability and frictional interaction between components. One of the aims of the Geosynthetics Lining System implemented is to separate the functions of the different items: stability is guaranteed by the geotextile of reinforcement (GtR) while the geomembrane (Gm) acts as the sealing layer and must be subjected to the minimum possible tensile stress. In addition a geospacer (GS) for transmissivity of water to drain is possibly inserted between the geomembrane and the geotextile of reinforcement. The (GtR) has also a filter function to avoid the clogging of the geospacer.

Different design methods based on simple limit equilibrium method are available (Giroud, 1989, Soyez et al, 1990, Koerner et al, 1991 (Figure 3), Poulain et al 2004). However, as demonstrated by a Finite Element Method approach (Villard et al, 1999), selected assumptions for the interface properties and boundary conditions are questionable.

Several unfortunate failures have resulted from soil sliding down the slick liner/drainage system interface. An accurate design of the (GtR) is required because this sheet is in charge of providing high frictional strength to the cover soil , but this condition is not sufficient: there is a high sensitivity of the relative displacements and mobilisation forces of the different components of the GLS to the interface friction relationship and the mode of construction.

In these conditions, case histories are needed to present a comprehensive view of this issue.

Anchorage

GLS stability

Erosion Control

118

Figure 3: Typical stability design method (Two Wedges method) for GLS (Koerner et al, 1991)

2.1 Tensile mobilization in geosynthetics during construction (Montreuil landfill experimentation):

The geosynthetic lining system (Figure 4), (Gourc et al,1997,Villard et al, 1999), supported by a clay base layer, consists of an HDPE geomembrane (Gm: J = 458 kN/m), a non-woven geotextile (GtR : J= 65 kN/m) and a cover soil 0.30 m thick granular soil layer. The friction angles g are 9 for Gm/support interface, 12 for GtR/Gm interface and 29 for the granular soil/GtR interface. The slope is 2H/1V.

Figure 4. Monitoring on a slope barrier of the Montreuil landfill

The forces acting on the geosynthetics at top of the slope were measured by force sensors positioned between the geotextile sheet clamps and the fastening posts anchored at the top of the slope. The displacements of the geosynthetics and the cover granular layer were monitored by means of cable-type displacement sensors linked to the fastening posts and regularly spaced on the geosynthetic sheets and in the granular soil layer.

The full experimental programme consisted of four successive implementation stage, but only the first one is presented here:

The layer of granular material is placed on the slope (up to a length along the slope and Lc = 6 m from the bottom). This experimental stage involves monitoring the forces and displacements in the various GLS components while loading the granular material layer metre by metre on the slope over a total loading length, Lc, of 6 m.

GtGm

Cable-type measuring

Granular material

Clay support

Fastening post

Toe stop

9 m2

1

119

Figure 5 illustrate the displacements in the granular material and in the geotextile during loading of the cover soil layer for increasing length (Lc). The displacements in the geotextile are far greater than those in the geomembrane. The relative downward displacement between geotextile and geomembrane induces a frictional force towards the bottom of the geomembrane. In agreement with the concept of separation of functions, the elongation due to tensile force in the geomembrane is very low. Figure 6 presents the evolution of the (extreme) tensile force in the geotextile and geomembrane sheets measured at the top of the slope with lifting of soil cover (Lc).

Figure 7 is especially interesting, seeing that it presents the strains in the geotextile sheet for increasing length (Lc). It is worth noting that the geotextile, acting as a reinforcement, is subjected to positive tensile strains (elongations) at the top of the slope whereas it is in compression at the bottom of the slope. Given that the slope length is constant, the elongation of the sheet at the top is compensated at the bottom by the formation of wrinkles (Figure 8).This complex behaviour is generally not taken in consideration in design and numerical calculations.

Geotextile

Granular soil

Figure 5. Montreuil landfill slope: displacements of the granular soil cover (down) and of the geotextile (up)

Figure 6. Montreuil landfill slope: tensile forces in the geotextile and geomembrane

0 1 2 3 4 5 6 7

0 1 2 3 4 5 6 7 8 9 L(m)

Lc = 6 m Lc = 5 m Lc = 4 m Lc = 3 m

Phase I Displacement (cm)

0 0.2 0.4 0.6 0.8

1 1.2 1.4 1.6

0 1 2 3 4 5 6 Lc (m)

Tension (kN/m) Phase I Geomembrane

Geotextile

0 1 2 3 4 5

0 1 2 3 4 5 6 L (m)

Displacement (mm) Lc = 6 m Lc = 5 m Lc = 4 m Lc = 3 m

Phase I

120

Figure 7. Montreuil landfill slope: Distribution of strains in the geotextile (elongation positive)

Figure 8. Mechanism inducing wrinkles at the base of the slope

2.2 Long term survey of a geosynthetic cap lining system (Experimentation of Torcy)

The site of Torcy is a landfill of domestic and non hazardous domestic waste. This experiment differs from the previous one by the far larger length of the slope (50 m instead of 9 m) , the inclination of the slope (3H/1V) , the long term monitoring of the movements of the GLS after construction (2 year) and the 4 different lining systems tested (Figure 9) (Villard et al,2000, Feki et al, 2002).

Figure 9. The 4 different geosynthetic lining systems on slope (Torcy)

Tension

Compression

GMb GS

GTr

Clay P2

GMpp

GTr

Clay GS

P1

Soil Cover Soil Cover

0.3

1 m GTs GS

GTr

Clay P3

Soil Cover

Clay P4

Soil Cover

-3

-2

-1

0

1

2

0 1 2 3 4 5 6 7 8 9

L (m)

Stra

in (%

)

Lc = 6m

Lc = 5m

Lc = 4m

121

The present presentation focus on (P1) and (P2) trials, which could be distinguished by the type of geomembrane: For (P1) it is a polypropylene geomembrane (Gm PP) and for (P2), it is a bituminuous one (Gm B). The differences in the mechanical characteristics are significant, related to the tensile stiffness (tensile modulus J) :

GtR ( J= 580 kN/m), GmB (J = 80 kN/m) , GmPP ( J = 51 kN/m) and the friction propertie g : Trial (P1): clay/GmPP (138) GmPP/GS (7) GS/GtR (18) Trial (P2): clay/GmB (18) GmB / GS (31) GS/GtR (18)

To attach the geosynthetics at the top of the slope, a large trench of anchorage (1 m deep and 1 m wide)

was used to bury the geosynthetics in such a way that no significative sliding of geosynthetics could be observed at the top edge.

Similar monitoring system than in Montreuil was used (metallic cables attached on every geosynthetic layer at different points), to evaluate the displacements along the slope. Short term behaviour (end of construction tc to tc+ 72h): SShort term

Short term behaviour just after completion of the cover soil on the geosynthetics was observed (Villard et al, 2000). Figures 10 and 11 show the distribution of the tangential displacement of the different components of (P1) and (P2) along the slope (distance L from the top ), between the time corresponding to the end of construction (tc) and time (tc + 72 hours): in both cases, a critical interface was noted depending on which the sliding (difference in displacement between the two geosynthetics in contact) was strongest: for (P1), between the geomembrane GmPP and geospacer GS (g = 7, interface with the least friction) and for (P2) between the geospacer GS and the geotextile for reinforcement GtR (g= 18, same condition).

Figure 10. Trial P1 - tangential displacements in the different components at tc

Figure 11. Trial P2 - tangential displacements in the different components at tc

0

50

100

150

200

250

300

350

400

0 10 20 30 40 50

GTrGSGMppClay

L (m)

Tang

entia

l disp

lace

men

tu

(mm

)

-100

102030405060708090

0 10 20 30 40 50

GTr

GS

GMb

Clay

L (m)Tang

entia

l dis

plac

emen

tu

(mm

)

122

The comparative behaviour of the two trials (P1) and (P2) clearly illustrates how it is possible to modify the whole deformation of a GLS only by changing the friction properties of one interface (here geomembrane/ geospacer).

Long term behaviour after construction (2 years):

The main studies available in the literature are limited to the short term behaviour of the geosynthetic liner on slope. For the present study, the two trials were monitored during two years.

Long-term monitoring is more complex, additional measures of the settlements of the waste body beneath the cap GLS should be needed because the waste embankment, consisting of domestic and non hazardous waste, is compressible and mechanically viscous: the profile of the slope is deformed, and more seriously, the monitoring table at the top of the slope settles, following the waste body deformation.

Figure 12, which plots the vertical settlement ( s ) versus elapsed time confirms the similarity of behaviour for the two profiles (P1 and P2) over time. The settlements are significative, due to the compressibility of waste. In these conditions the GLS should accommodate and conform to its support without ruin of its functional properties.

The results collected from this case historie are remarkable since for the two trials, the tangential displacements of all the components of the GLS are of the same order than those in the cover soil (Figure 13). So there is practically no additional relative displacement between the different sheets after the construction stage (tc+ 72h).So no additional tensile mobilization in the GLS can be expected.

Figure 12. Torcy landfill: Vertical settlements along the slope (L= 5m and L = 30 m from the top )

0

100

200

300

400

500

600

0 3 6 9 12 15 18 21 24

P1 (L=30m)P2 (L=30m)P3 (L=30m)P1 (L=5m)P2 (L=5m)P3 (L=5m)

(t-tc) months

settl

emen

t (m

m)

tc(end of construction) + 72h

L

L= 5m L=30 m

s

s

123

Figure 13. Torcy landfill: Evolution during 2 years of the tangential displacements along the slope for the different components of the GLS (Trials P1 & P2)

2.3 Conclusions

The observation of the actual behaviour of GLS on slope demonstrates that their mechanical behaviour is complex and really difficult to modelize: extreme sensitivity to the interface friction properties, influence of the construction conditions, influence of the compressibility of the waste body for cap liners.

In the next Chapter 3, assessment of interfaces friction will be considered and in chapter 4 ,a decisive boundary condition for the GLS, the anchorage strength of the geosynthetics at the top of the slope will be analysed.

3 USE OF INCLINED PLANE TO ASSESS STRESS MOBILIZATION OF LINER ON SLOPE

Figure14. Assessment of the interface properties of the different components of the GLS

0

50

100

150

200

0 3 6 9 12 15 18 21 24

cover soilGTrGSGMppclay

(t -to) months

Tang

entia

l disp

lace

men

tu

(mm

)

0

50

100

150

200

0 3 6 9 12 15 18 21 24

cover soil GTr GS GMb clay

( t - to ) months

Tang

entia

l dis

plac

emen

t u

(mm

)

GLS stability

124

The large scale experimentations presented above have demonstrated behavioral sensitivity to small

modifications in, for example, the friction interface relationship. The complete friction interface relationship (shear stress vs. tangential displacement and, ultimately, vs. time t) often proves necessary in explaining distribution of tensile forces and relative displacements in the Geosynthetic Lining System. Hence, thorough understanding of the complete friction interface relationship ( /' = f( , t) at a fixed normal stress '), and not just friction limit values (sg and gg with limit /' = tan , threshold value) of the soil-geosynthetic and geosynthetic-geosynthetic, is required to the stability analysis of sloped systems (Gourc et al ,2004).

These interface relationships are determined using devices of either the shear box or inclined plane type and such equipment is currently undergoing standardization (European Standard final draft prEN ISO 12957, 2001: Article 1 for the Shear Box test, Article 2 for the Inclined Plane test , Gourc et al , 1996).

The inclined plane test is commonly used when studying the stability of sloping geosynthetic liner systems under conditions of low normal stress. The inclined plane offers the dual advantage of enabling testing at low normal stresses at the interface and allowing for test condition modulation. The rational minimum normal stress is = 25 kPa for the Shear Box test which is higher than the actual stresses induced by a layer of cap soil cover .So Inclined Plane test is generally preferred for the design of cap liner.

Both the shear box (SB) and the inclined plane (IP) tests are presented on Figures 15 and 16 respectively. Shear box device is initially based on a large direct shear equipment. In addition the present box has been

adapted to fit the need of a uniform distribution of the normal stress ', thanks to an hydraulic bag set under the compression plate.

Inclined plane device is specially designed for tests on soil- geosynthetic or geosynthetic-geosynthetic interfaces. This is the adaptation of the two devices to this configuration which is presented on Figures 15 and 16. In addition, adaptation of the device for simulation of water flow at the interface is also possible (Gourc et al,2001, Brianon et al , 2002) .

Figure 15. Shear Box Device ( SB ) .Adaptation to Geosynthetic / Geosynthetic interfaces

Figure 16. Inclined Plane device ( IP ): adaptation to geosynthetic / geosynthetic interfaces

(t)

w (t)

Upper geosynthetics

Lower geosynthetics

Displacement monitoring

Collect of data

Tilting system

125

0

5

10

15

20

25

0 5 10 15 20 25 30 35 40

(mm)

(kP

a)

25

50

75 ' (kPa)

3.1 Comparison between IP and SB results:

(Lalarakotoson et al, 1999) demonstrated that the SB test results are globally consistents with the IP tests, as well as the modification of shear behaviour with the normal stress is taken into account.

The results for IP and SB tests for interface between a reinforcement geotextile (GTr) and a geospacer (GS), are presented (Reyes et al, 2002 and 2003):Comparison of threshold angle of friction (gg) values:

The SB tests (Figure 17) submitted to three normal stresses (' = 25, 50 and 75 kPa) gave for gg a value of 16.8. It should be noted that no peak frictional strength is observed for this particular interface. For the IP test (Figure 18) with '0 = 5.7 kPa, yields at s = gg = 18.4. This drop in the angle of friction with ' increasing is often considered as similar to the phenomenon well known for granular soil , but ,as demonstrated in 3.4 it could be also ascribed partially to an overestimation of gg due to a standard interpretation of the Inclined Plane results.

The following must nonetheless be highlighted: Test kinematics are considerably different in the direct shear box and the inclined plane tests. No strain

softening has been observed within the shear box tests; however, residual friction cannot be measured under any circumstances with the inclined plane given that non-stabilized sliding arises at constant /' values (it would be necessary, to demonstrate strain softening, to reduce inclination once instability is first detected!).

Using the inclined plane, the threshold angle stipulated by European Standards is typically obtained for the soil-geosynthetic test with a relative displacement = 50 mm, which remains extremely high, at least for geosynthetic-geosynthetic interfaces. Furthermore, behavioral information available on the phase preceding non-stabilized sliding, which may be quite distinct from one geosynthetic to the next, is not taken into account. Behavior during this phase could serve to distinguish between various interfaces all displaying the same gg value (Figure 19). This observation instigates a comparison of the behavior during the phase of small displacements prior to the threshold sliding phase characterized by angle gg.

The effective normal stress ' decreases throughout the inclined plane test, and the parameter/' = tan

was selected to allow comparison with the shear box test. Displacement and stress ratio /' relationships for shear box and inclined plane tests are shown in Figure 20 from Figures 17 and 18 for the same GTr-GS interface. The results presented are for normal stresses ' = 25, 50 and 75 kPa for the direct shear box, while ' varies in the inclined plane test (' = 5.4 kPa for s = 18.4 at = 50 mm ). The displacement corresponding to maximum friction (s) appears considerably smaller within the inclined plane test, with this observation not necessarily due to the lower normal stress. The shear box test set-up does not enable testing under very low normal stress conditions (' = 5.4 kPa)

.Knowing the displacement (s) is essential when seeking to incorporate an accurate interface friction relationship for an elaborate computation of geosynthetic liner system deformation on slope, e.g. using the finite element method.

Figure 17. Shear Box test (GTr-GS): shear stress ( ) versus displacement ( ) for 3 different normal stresses (' ).

126

Figure 18. Inclined Plane test (GTr-GS): displacement ( ) versus slope angle ( ) for ('0 = 5.7 kPa).

Figure 19. Different patterns of the friction behaviour in SB and IP tests

Figure 20. Attempt to compare (IP) and (SB) test on the same diagram: (GTr-GS interface).

0

20

40

60

80

100

0 2 4 6 8 10 12 14 16 18 20

(

(m

m)

gg = 18.4

/

tan gg

gg

SB IP

s

s

0

10

20

30

40

50

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4

/ '

(m

m)

SB

IP

50 7525

127

3.2 (IP) test to find the optimal position of the Geosynthetic reinforcing the soil cover (GtR)

The Inclined Plane device is specifically adapted to highlight the role of reinforcement of a geosynthetic, stabilizing the soil cover ( Lalarakotoson et al , 1998 ).

A schematic side view of the test apparatus is shown in Figure 21 The reference case, ie unreinforced, includes a compacted soil layer filling an orthogonal rigid box , a

geomembrane placed on a rigid horizontal PVC support and connected to an anchorage device by clamps. The PVC support is fixed on the tilting base plane ( geomembrane/support = 20) . The upper box is supported on the base plane by means of metallic rollers on both sides.

In the present case, the reinforcement consists in placing a geosynthetic at different positions (d) in relation with the thickness of the soil (Figure 22). It is worth noting that a similar case was recently studied by ( Palmeira et al, 2003).

The system is mobilized by tilting progressively the plane at low constant vertical rate. The global state of the system can be described by the angle of plane inclination , the relative displacement u of the box and the maximal axial force T mobilized within the different geosynthetics.

The diagramme (Figure 23) displays the complex behaviour of the system and the evolution of the distribution of forces between the geomembrane (TGm ) and the geosynthetic of reinforcement (TGtR) and the evolution of the displacement (u) of the soil cover. The reference is the case without reinforcement of the soil cover.

The (GtR) is a geogrid (J= 650 kN/m), positioned in three different position (Figure 22), close to the geomembrane (traditional position:base) or at d = 0.05 m or 0.15m above the geomembrane base. The geomembrane has a stiffness (J= 450 kN/m).The soil cover is a dense sand ( peak = 39 , res = 32).

The benefit gained including a reinforcement of the soil layer is clear: Sliding of the upper box is obtained for an inclination of the plane only = 27 without GtR.In this

experimental configuration, friction to guidance of the upper box should be subtracted to obtain gs geomembrane-soil = 28.4 at peak) . On the other hand the efficiency of the reinforcement geosynthetic is decreasing when the geosynthetic is closer to the geomembrane: tensile force in the geomembrane TGm higher and displacement (u) higher for the same inclination .

However, due to field difficulties to lay the geosynthetic of reinforcement into the thickness of the soil cover in place of directly on the geomembrane base, the traditional position of the GtR on landfill slopes remains the position on contact with the geomembrane.

Figure 21. Inclined Plane test: Adaptation to the evaluation of the reinforcement of a soil layer

Boitier (100 x 70 x 30 cm3)

Renfort

Gomembrane

TRenfort

TGm

15

5

10 cm

1

PVC support

Capteurs de tension

LVDT

40

128

Figure 22. Different positions of the geosynthetic of reinforcement in the IP upper box

Figure 23. Relative displacement (u), Tensile force in the Geosynthetic of reinforcement and in the Geomembrane, due to the inclination of the plane, for different positions of the reinforcement.

3.3 (IP) to emphasize the creep behaviour of some geosynthetics interfaces:

The inclined plane test, in comparison with the shear box test, features the possibility of easily conducting tests at constant stresses (',), a situation that corresponds with the actual case on slope in the field.

Tests of this type, called "creep tests", were performed at the Lirigm on interface between Geomembrane and Geonet used at the base of the Geosynthetic Lining System (Reyes et al , 2003).These tests consist of raising the inclined plane (at a constant d/dt), i.e. the upper plate with its geomembrane in a fixed position ( = 0) up to an angle of less than the non-stabilized sliding value ( = c < s), with s the inclination corresponding to the non-stabilized sliding .The next step calls for observing the evolution in displacement with constant = c while freeing the upper plate.

Once a plane inclination of c has been reached, the upper plate is released and sliding displacement (c) is recorded vs. time (t - tc) at constant plane inclination c.

The case of a polypropylene geomembrane GMpp is examined, since this exhibits more gradual sliding during the standard test at constant inclination rate: Figure 24 plots the results of a standard test, with displacement () versus time (t) instead of inclination (). Given a rate d/dt = 3/min, the displacement can no longer be related as a function of but rather of time t, for (c) values sharply lower than (s) (corresponding to the non-stabilized sliding). Displacement () in the standard test is indeed non-negligible ( = 2 mm for = 10.2 < s = 14.2).

0

5

10

0 5 10 15 20 25 30 35 40 45

u (

mm

R encebasebase + 5 cmbase + 15 cm

no reinf.

GtRd = 0

d = 0,05 d = 0,15

Cover Soil

- - - - - without reinforcement base: d = 0 d = 0,05 m d = 0,15mGtR

GtR

0,30

m

0,05

m

0 ,15

m

Base Gm

0

0.2

0.4

0.6

0.8

0 5 10 15 20 25 30 35 40 45

T G

m (

kN

)

R encebasebase + 5 cmbase + 15 cm

- - - - - without reinforcement base: d = 0 d = 0,05 m d = 0,15m

GtR

Geomembrane

no reinf.

d = 0,05

d = 0,05 d = 0,15

129

Static Test GMpp-GS (t & ) ' 0 = 5.7 kPa

0

2

4

6

8

10

40 60 80 100 120 140 160 180 200

t (sec)

(m

m)

= 9

= 11

= 10

d /dt =3min

s = 14.2

Figure 25 shows the creep results obtained for three different inclinations c: even for c values much lower than the threshold inclination s, the displacement is not stabilized. This important finding may lead to reconsidering the significance and technique for determining the threshold angle of friction gg in the standard test since non-stabilized sliding is obtained for c < s. It may also be observed that the creep displacement rate d/dt at identical time (t tc) increases when nearing s.

The representation of the standard test diagram in Figure 24 may be completed in order to include the creep phase. Figure 26 combines the standard test displayed in Figure 24 with the creep test in Figure 25.

The results in Figure 26 show what the sliding behaviour of the interface would be if a standard test were interrupted prior to reaching the non-stabilized sliding phase.

The same tests were conducted for a Geomembrane HDPE in contact with the same Geospacer, test on GMhdpe-GS interface and it was found that even if a zone of gradual sliding (less pronounced than for GMpp) is present within certain standard tests, the tests carried out with various c values do not show any creep sliding even after being run for 12 hours. Displacement only occurs once inclination has equaled or surpassed s. In Figure 27, we combined (as we did for GMpp) the results of a standard test (inclination vs. time) with those of a creep test for the GMhdpe-GS interface.

In this particular case, creep at the interface is not highlighted as was the case for the GMpp-GS interface. Creep behavior therefore does not systematically appear for all geosynthetic interfaces.According to this serie of tests and in the case of interfaces exhibiting gradual sliding "strain hardening" prior to the non-stabilized sliding phase (i.e. high s), creep for c < s (the case of GMpp) is still observed whereas for interfaces displaying more sudden "brittle" sliding (the case of GMhdpe), creep is no longer observed. This statement should however be confirmed on other interfaces.

Figure 24. Standard Inclined Plane test on GMpp-GS interface: displacement ( ) versus time (t) instead of inclination b ('0 = 5.7 kPa).

Figure 25. Creep test on GMpp-GS interface, for three different constant inclinations c.

Creep Test GMpp-GS ' 0 = 5.7 kPa

0

20

40

60

80

100

0 1 2 3 4 5

t-t c (hours)

(m

m)

c = 9

c = 11 c = 10

t c

130

0

1

23

4

5

6

78

9

10

120 140 160 180 200 220 240 260 280

t (sec)

(m

m)

d /dt = 3min

c =

c = 13,8

c = 12 c = 11

c =

(standard test)

(creep test)

Figure 26. Inclined Plane test on GMpp-GS interface: combination of a standard test diagram with three creep test diagrams (inclination c). Figure 27. Inclined Plane test on GMhdpe-GS interface: combination of a standard test diagram with five creep test diagrams (inclination c). 3.4 Interpretation of the dynamic phase during an Inclined Plane test

This is a new and more comprehensive interpretation of the Inclined Plane test.

Indeed in the present time only the limit inclination = S corresponding to the unstabilized sliding along the slope is considered. With a negligible friction due to guidance, the ISO standard 12957-2 gives: tan s = tan S conventionaly evaluated for a sliding displacement = 50 mm. A more careful study could be fruitful as demonstrated in the presentation below. The behavior may be separated into several phases ( Figure 28 ), as follows: Phase 1 (static phase): upper box practically immobile ( = 0) over the inclined plane until reaching an

angle = 0; Phase 2 (transitory phase): for an increasing rate of inclination ( > 0), upper box moving gradually

downward direction. Phase 3 (non-stabilized sliding phase): upper box undergoing non-stabilized sliding at an increasing speed

(d/dt), even if plane inclination is held constant ( = s).

0

1

2

3

4

5

6

7

8

9

10

40 60 80 100 120 140 160 180 200

t (sec)

(m

m) c = 11

c = 10 c = 9

d /dt =3min s = 14.2

(standard test)(creep tests)

131

Figure 28. Different phases of the upper box movement, for increasing inclination of the inclined plane.

In the literature, nobody pays attention to the different behaviour of geosynthetics during this phase. As demonstrated in (Gourc et al, 2004) ,the analysis of this phase could be instructive.

From the inclination value = s, the sliding rate of the upper box becomes significant and the mechanical analysis must absolutely be conducted in dynamics (taking into account the displacement acceleration ) and no longer in statics, as is typical practice.

In Phase (3), = s and it was demonstrated (Gourc et al,2004) that the upper box is in a state of movement, with a constant acceleration r c .It is possible to determine a new friction angle:

gc

ss

dync

= cos

1tan tan . Finally it is possible to distinguish three friction angles: - tan stats = tan s : standard but non accurate ( and non conservative) friction angle corresponding to the

unstabilized sliding movement - tan stat0 = tan o for a displacement = 0 : conservative friction angle corresponding to the initialization

of the movement - tan dync calculated in dynamic conditions for movement with constant acceleration Figure 29. Evolution of the displacement , the acceleration and the interface friction angle with the plane

inclination (sudden sliding) The Figure 29 presents the corresponding schematic evolution during an IP test of the displacement , the acceleration and the actual friction angle with the inclination . As indicated on Figure 30, in fact Phase 2 may be of various types: Type a) sudden sliding - abrupt displacement of the upper box under non-stabilized sliding with a nearly-

inexistent Phase 2 (0 = s); Type b) jerky sliding - displacement () increasing in a "stick-slip" fashion; Type c) gradual sliding - displacement () progressively increasing with inclination ().

0 = s

dync

stat0

c

132

It was demonstrated a correlation between the different friction angles depending of the type of phase 2: Sudden sliding(Type (a) Figure 30 is observed for (tan stat0 > tan dync ) Gradual sliding (Type (c) Figure 30) is observed in opposite case for tan stat0 < tan dync .Friction is increasing with the displacement () and the upper box sliding requires an increase of the inclination from o to s to reach the non-stabilized sliding Figure 30. Different mechanisms of sliding observed in the Inclined Plane test: (a) sudden sliding (b) jerky sliding (c) gradual sliding

For instance, lets consider the test on interface Geomembrane HDPE / Geospacer .A test of the same set is presented on Figure 27. On the figure 31, a comprehensive dynamic interpretation of the test is presented. A fine analysis of the movement phase allows the determination of a phase of uniform acceleration (displacement rate proportional to time): Initialization of the movement is obtained for : tan o = tan stat0 = tan 10.0 Taking into account the acceleration c = 0.35 m/s2 the formula above guives : tan dync = tan 11.9 The conventional "static" interpretation sharply overestimates the interface friction : tan stats = tan s = tan 13.9 Finally the observation of a gradual sliding (type (c) Figure 30) is fully compatible with the friction results, seeing that tan stat0 < tan dync (friction increasing with movement).

(mm)

() (1)

(2)

(3)

0 s

50 mm

(mm)

() 0 s

(mm)

() s Type (a) : sudden sliding

Type (b) : jerky sliding Type (c) : gradual sliding

133

Figure 31: Comprehensive dynamic interpretation of an interface friction test at the Inclined Plane: (1) Displacement / inclination - (2) Inclination / time t (with = 0 for t = 0 ) - (3) Displacement / time t and Displacement rate / time t during the non-stabilized sliding phase ( Figure 28 ) .

GMhdpe-GS

0

0,02

0,04

0,06

0,08

0,1

0 2 4 6 8 10 12 14 16

(

(m)

0

50

GMhdpe-GS

0

0,02

0,04

0,06

0,08

0,1

160 180 200 220 240 260 280 300

t (sec)

(m)

0

50

GMhdpe-GS ( '0 = 5,7 kPa)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

262 263 264 265 266 267 268

t (sec)

(m)

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

(m

/s

displacement displacement rate

ti

tmax

134

Finally the conventional friction angle stats is not theoretically justified and it is systematically greater than

stat0 and generally greater than dync . Consequently using conventional friction angle stats for assessing stability of geosynthetic liners in slopes is not conservative. It would be reasonable to revisit the ISO-CEN 12957-2 corresponding standard. The decision to select either stat0 or dync for design of geosynthetic systems on slopes can be influenced by a new interpretation of the creep tests already displayed above (Reyes and Gourc, 2003). 4 ANCHORAGE CAPACITY OF GEOSYNTHETICS IN TRENCHES

Figure 32. Evaluation of the pull-out strength of the geosynthetic anchorage

The behavior of the anchorage of geosynthetics sheets at the top of a slope is a decisive factor when it comes to determining the dimensions of geosynthetic lining systems on slopes. In order to optimize the geometry of the structures in question (to reduce the area taken up by the anchorage at the top of the slope), anchorage solutions using trenches of varying forms are sometimes used. Calculating the required dimensions of this anchorage remains problematic. In order to improve knowledge of the behavior of anchor trenches, experimental and numerical studies at the Lirigm in Grenoble (Chareyre et al,2002 and 2004 ) and at the Cemagref in Bordeaux (Brianon et al, 2000) were carried out .

The role of the anchor is to withstand the tensile force generated by friction along the slope (Figure32) as evaluated on site in the Chapter 2. This can be done using a simple run-out (linear) anchor . However, the geosynthetic sheets are often installed in trenches, with a L-shape, V-shape or U-shape, to optimise the dimensions of the anchor zone (minimal horizontal area occupied) and to ensure effective anchorage. To size the system, it is necessary to estimate the tension that can be mobilised in the anchor (the anchoring capacity) according to its geometry and the properties of the constituent materials. Very few large scale experimentations were carried out (Imaizumi et al,1997) in order to calibrate the traditional design methods . The target of the present work is to improve the design, taking into account the complex interactive behaviour of soil and geosynthetic anchor.

4.1 Pull-out experimentations

Small scale experimentations Firstly, preliminary anchorage tests were carried out at the Lirigm on a small scale model in an analogic

two-dimensional granular soil (Schneebelli rollers composed of 5mm and 3mm diameter and 60 mm length duralumin rollers, friction angle = 22). The sheet was a thin non woven geotextile (Villard et al, 1997). The stereophotogrammetric technics was used to get the displacements field (Figure 33).

Several trial tests were performed, comparing the pull-out behaviour of sheets of similar length (L= 0.30 m or 0.40 m) but different geometry. In each test, the height of the soil above was H= 0.15 m.

(b

(c)

(d

Anchorage

Friction

(a)

Soil layer

135

On Figure 34, the corresponding pull-out force is reported versus the displacement (U0) of the anchor head. It is worth noting that, for the same sheet length, the anchorage geometry has a little effect on the anchor strength, but the maximum pull-out force is reached for a larger displacement for the U-shaped anchor.

Figure 33 recording the displacements for soil and sheet highlights the mechanisms involved during the pull-out process:

For the linear anchor , the failure of the soil above the sheet is clearly corresponding to the assumption of failure of the soil mass above the sheet in accordance with one of the analytical models proposed below (Figure 39 : displacement of the block (A)) .

For the L shaped and U-shaped anchors, compression of soil at the bends (Figure 35). This is a significant mechanism, difficult to take into account in a theoretical approach.

Figure 33. Field of displacement in the geotextile and in the soil around (analogic Schneebelli soil)

during pull-out test for anchorages of different shapes ( [1] linear, [2] L-shape and [3] U-shape) .

Figure 34 : Evolution of the pull-out force for anchorages of different shapes (analogic Schneebelli soil).

(1)

(2)

(3)

0

20

40

60

80

100

0 20 40 60 80 100 120

Head displacement (cm)

Hea

d fo

rce

(N)

U-trench Ln=40cmlinear Ln=40cmL-trench Ln=30cmlinear Ln=30cm

136

Figure 35 : Typical mechanism of compression at the anchorage bend, during pull-out test.

Large scale experimentations

The main large scale experimentations were carried out at the Cemagref of Bordeaux (Brianon et al, 2000) on an anchorage bench of large dimensions (Figure 36) offering the possibility of performing full-scale pull-out tests. Tests with specific instrumentation were carried out on sandy silt and on sand, in order to provide a better understanding of the phenomena involved and so as to be able to compare the results with those from numerical modeling ( Chareyre et al, 2002 and 2004, Villard et al , 2004 ) with on one hand an analytical approach and on the other hand the Dictinct Element Method ( DEM).

Figure 36. Large scale experimentation related to geosynthetics pull-out tests ( Cemagref-Brianon et al,2000).

The anchorage bench was composed of an anchor block of a width of one meter and a traction system. The dimensions of the anchorage zone allowed for an anchor trench with a total depth (D+H) equal to 1 m and a length covered (L+B) of up to 1.2 m. The traction system was composed of a winch with a maximum capacity of 50 kN and a pulley that allowed us to exert the tensile forces on the geosynthetic at incline angles of between 0 and 35. This traction system was fixed onto the geosynthetic using a metal clamp. An important specificity of this model is the direction parallel to the liner slope of the tensile force at the anchorage head. The geotextile remained in contact with the slope, due to a small deformation of the soil in the early stage of the test.

To limit edge effects, the side walls of the anchorage bench were covered with a smooth, polypropylene geomembrane. The sand / geomembrane friction angle was about 20. The soil was installed in successive layers and then compacted.

The tensile force T and the displacement U0 of the traction cable were measured on pulling out via sensors fixed onto the traction system . In the anchorage zone , a cable measuring system was used to monitor the displacements of the geosynthetic at different points . The cables could slide through the flexible sleeves isolating them from the soil, were fixed to the geotextile sheet, and were tensioned using counterweights.

In certain cases, the movement of the soil could be observed thanks to columns of colored sand positioned in the anchorage zone before starting the test (Figure 37). After the geosynthetic sheet had been pulled out, the area of the soil directly above the sand columns was meticulously cut into sections to analyze the failure mechanisms and the displacements of the soil.

Anchor Block Traction system

Geosyntheti

Soil

Pulley

Winch

Metal clamp

1 m

H D

B L

1.5 m

Side wall

137

A frictional cohesive soil (Soil 1 silt: = 34 and c= 11,7 kPa ) and a purely frictional soil (Soil 2 sand := 41 and c= 0) were used .The geosynthetic used for the tests was a non-woven reinforced in one direction. It had a tensile stiffness modulus J of 624 kN/m (measured at 12% of strain), and the soilgeosynthetic friction angle g was equal to 34 with both soils.

Only the results of the anchor trench with the L-shape carried out with the sand and the silt are presented here (Figure 42 ).

(a) Sand (b) Sandy silt Figure 37. Photos featuring soil displacements, thanks to colored sand, close to a L-shape anchorage bend ,

for two different soils (Cemagref experimentation) .

4.2 Existing design methods

Researchers who have proposed analytical formulas for designing run-out or trench-type anchors ( Koerner

1998; Guide technique 2000) drew on various hypotheses for determining the maximum force Tmax that can be mobilised in the anchor. Certain hypotheses are common to all the authors, while others vary from one to another. These hypotheses are summarised below:

Assumptions common to all researchers

- The geometry of the anchor is represented schematically by linear segments (numbered in an increasing order from the outside towards the inside of the soil mass).

- The anchor fails only by relative displacement at the soil/geosynthetic interface. - The shear stresses that can be mobilised at the interface are equal to the maximum stresses max

corresponding to the slip limit state (on one or both sides of the geosynthetic). - Friction is governed by a Mohr-Coulomb interface law: max = n tan g with g and n the friction angle

and normal stress acting at the interface before pull-out. - The contribution of the ith segment to the total anchorage can be assimilated to a force Fi calculated by

integrating the shear stress of intensity max on either side of that portion of geosynthetic sheet.

Assumptions specific to individual researchers

- Koerner (1998) takes into account a different stress state on the two sides of the L-shaped anchor (Figure 38-Method A).

138

The inclination of the tensile force T1 produces an increase in the normal stress under the first segment of the anchor. This increase is assumed to be equivalent to the vertical component of T1. This means increasing the tension that can be mobilised by a factor of 1/[1sin()tan(g)], where is the slope angle and g the friction coefficient at the interface. On either side of the vertical segments, the normal stress is the active or passive earth pressure, depending on the side considered. If it is extrapolated for a 3-segment anchor (L-shape):

( ) ( )[ ] tansin131max = ==ii iFT (Method A) - In Guide technique (2000), as shown in ( Figure 38-Method B), it is considered that the tension in the

sheet is multiplied at each change in direction by a factor of e tan g ( being the difference in orientation between two segments in radians). This exponential factor is derived from an analogy with the friction of a wire on a cylinder. Implicitly, this means taking into account an increase in the normal stress n at each bend (refer to the mechanism of Figure 35. Outside the bends, the normal stress at the interface corresponds to the initial stress state. Finally, for three segments, with Ki being the coefficient associated with the bend i, one obtains:

( )[ ]332211max FKFKFKT ++= (Method B)

It has been learnt from various case studies that some of these hypotheses are only appropriate under certain conditions. In particular, relative displacements at the interface do not always occur, the friction at the soil/geosynthetic interface may be only partially mobilised if the failure occurs in the soil, and the normal stresses acting on the interfaces at failure can be very different from the initial stresses.

Figure 38. Different hypotheses for normal stress state at the interface for L-shape anchor:

Method (A) Koerner (1998), Method (B) Guide Technique (2000).

T1

Kp (H+D)

Kp H

H

Ka

Ka

(method A)

H H 2 T1 sin() / T1 sin()

T2

T2

T1 T2 T1 T1

K0 (H+D)

K0 H

K0

K0 H (method B)

H K1 = e tan = T1 / T

H

D

139

4.3 New analytical model for L-shaped anchors

Consider an L-shaped anchor in a rectangular trench as shown in Figure 38 , Villard et al ,2004 proposes a new analytical model ,taking into account the complex mechanisms involved during the pull-out of geosynthetics sheets. In addition (Villard et al , 2004) proposed a numerical modelling of the same problem based on the Distinct Element Method (D.E.M.) , but it is not presented here.

Figure 39.Notation corresponding to the new analytical model for the anchorage pull-out strength

Lets consider firstly the linear segment (1-2) of the sheet (Figure 39):

The tension T1 is mobilised by friction at the soil/geosynthetic interfaces while the upper layer (Block A) tends to move at the same time as the sheet.

The static equilibrium of Block A above the sheet can be used to determine the maximum value of T1 . P1 is the weight of the soil above the sheet and Rt the tensile force mobilised in the soil at the end of the anchor. and t are respectively the unit weight and limit tensile stress in the soil. The forces are expressed per unit width of the anchor. It should be noted that the strength of the upper layer may be greater than P1tan(g). In this case, it is assumed that the cover soil remains integral with the soil mass and the force Rt that can be mobilised is equal to P1tan(g), in conformity with equation [2]. [1] T1 = P1 tan(g)+ Rt [2] Rt = min[H t , P1 tan(g) The tension T1 required to pull out the sheet (applied parallel to the slope) is obtained by considering the angle effect. Two failure mechanisms are considered:

T3 T3 T3

T2

T2

T1 T2 T1 T1

T3

T2 T3

P2

T2

Fh

T3

L

H

T1

B

D

Block A Block

B 3

2 1

140

(A1) Assumption of a rigid soil mass

It is assumed that near the change in direction, the geosynthetic moves in relation to the underlying soil, which is considered to be fixed. It is therefore assumed that the stress state at the interface verifies = n tan(g). In this hypothesis, the problem is analogous to that of a wire rubbing against the arc of a circle(Figure 35). It is thus possible to introduce a weighting coefficient K1, the value of which is given in equation [3], such that K1 = T1 / T1. It should be noted that this result is obtained by applying the approach recommended in the Guide technique (2000) to a particular case.

[3] K1 = e tang

(A2) Assumption of soil mass failure

In this case, it is assumed that the forces acting near the change of angle are likely to shear the soil along the slip line shown in Figure 40, and that the moving block is sufficiently small for the volume (weight) and surface (surrounding stress) forces applied to the block to be neglected in relation to T1 and T1. Then, considering Mohr-Coulombs law of friction along the slipe line, a limit equilibrium of the forces in the sheared zone yields equations [4] and [5] (with being the internal friction angle of the soil) :

Figure 40. Assumption of a failure of soil near an anchorage bend [4] T1 T1 cos() R sin( + ) = 0 [5] T1 sin() + R cos( + ) = 0

The most critical angle (the one that minimises the ratio T1 / T1) is zero. The possibility of failure within the soil mass does not exclude failure at the interface. The predominant mechanism is that which leads to the lowest ratio. To summarize, the weighting coefficient K1 can be used to take account the effect of change of direction in the following way:

for a rigid mass:

[6] T1 = K1 T1 with K1 = e tan g for a mass failure: [7] T1 = K1* T1 with K1* = min [e tan g ; cos() + sin() tan( ) ]

R T

T

X

Y

141

By combining equation [1] with equations [6] or [7], depending on the assumptions made, it is possible to determine the anchoring capacity for the linear horizontal segment (1-2).

Anchors with multiple changes in orientation Similar coefficients to those in equations [6] and [7] will be derived associated with each bend. The

variables incorporating this possibility of failure in the soil is indicated with an asterisk: RL, the sum of the friction values on the upper side of segment (1-2), is a function of the geometry and characteristics of the soil. It is considered that : - If Ht > P1tan, the soil cover (Block A) is sufficiently strong not to be dissociated from the rest of the

soil mass (Block B). During pull-out, it is assumed that RL = P1 tan. - If Ht < P1tan, Block A will be dissociated from the rest of the soil mass after cracking at bend 2.

Given the fact that the anchor is mobilised progressively as the sheet pulled out, the failure of the upper block of soil will occur before that of the anchor. In this case, only friction at the base of the soil layer will be considered in determining the anchoring capacity. In this case, it is assumed that RL = 0.

Two failure mechanisms are considered:

(A1) Assumption of a rigid soil mass

In the case of a rigid soil mass, failure involves the relative displacement of the inclusion in relation to the mass, through slip at the interface. Thus the maximum forces that can be mobilised correspond to the limit equilibrium state at all points of the soil/geosynthetic interface (i.e. = n tan g). The value of the anchoring capacity is therefore determined by considering the distribution of the normal stress n on each segment of the interface. - Segment 3: At the initial state it is assumed that the weight P2 of Block B of the soil mass rests entirely on

the last segment of the sheet. When the tension in the sheet increases, the action of the geosynthetic at bend 3 results an uplift force on Block B. The vertical component of this uplift force is equal to T3 (Figure 39), and opposes to weight P2. It results in a reduction in the normal stresses acting on the segment 3 of the anchor, and the final normal stresses is equal to (P2 T3). Hence, by using the notations in Figure 39, T3 is given in equation [8}, and the relation between T3 and T3 at the limit slip state is defined in equation [9].

[8] T3 = 2 (P2 T3 ) tan g

[9] T3 = K3 T3

Where P2 is the weight of the soil above the sheet ( P2 = B (H+D) ) and K3 is the change-of-angle coefficient defined as in equation (11): K3 = e /2 tan g.

Thus gives:

[10] T3 = 2 P2 K3 tan() / [1 + 2 K3 tan() ]

- Segment (2-3): Fh is used to denote the horizontal forces exerted by the soil on the vertical portion of the sheet. Fh is obtained by equation [11]. The friction force (proportional to Fh) that can be mobilised on each vertical side is added to T3 to give the tension T2 [12]

[11] Fh = Ko D (H+D/2)

[12] T2 = T3 + 2 Fh tan(g)

[13] T2 = K2 T2

Ko is the earth pressure coefficient at-rest.K2 is the change-of-angle coefficient defined by: K2 = e /2 tan g

142

- Segment (1-2): The proposed mechanism for this part is the same as for the run-out anchor, with Rt = RL T1

[14] T1 = T2 + P1 tan(g) + RL [15] T1 = K1 T1

K1 is the change-of-angle coefficient defined by: K1 = e tan g.

(A2) Assumption of soil mass failure

For the upper bend, the possibility of failure is envisaged by taking the sub-horizontal failure line of Figure 40 as simplifying assumption. This failure scheme, which is similar to that adopted for the failure of a soil bend but extended to a larger area, was derived from the results of simulations (Chareyre 2003). The stability of Block C situated between the geosynthetic and the failure line is considered. Complete formulation of tensile pull-out force in the anchor is presented in (Villard et al, 2004).

4. 4 Interpretation of the large scale experimentations

Comparison of the performed tests is specifically interesting because the interface friction ( g ) is the same for the two soils (silt and sand) despite their different intrinsic mechanical properties.The results obtained show that the anchorage capacities are much greater with the sandy silt, despite the similar interface characteristics .These results show that the soil plays a major role in anchorage mechanisms, and that it is not enough to take into consideration only the interface friction characteristics when determining anchorage capacities.

Figure 41 : Simulation of the observed mechanisms during the pull-out experimentation, for cohesive (a) and cohesionless (b) soils

(a)

(b)

(c)

initial state

Silt

Sand

143

Figure (41) is a simulation by Distinct Element Method of the observations at the end of the two tests after pull-out of the sheet in the case of the cohesive frictional soil and in the case of the purely frictional soil respectively. With Silt, there is no major deformation of the soil mass. With Sand, in contrast, there is severe deformation. This confirms the pertinence of two different mechanisms : for the present conditions, the "rigid soil mass" assumption (A1) is appropriate for the cohesive frictional soil and the "soil mass failure" assumption (A2) is appropriate for the purely frictional soil.

The analytical formulations developed previously in ( 4.3 ) are compared in Figure 42 with the available experimental data ( assumption of a rigid soil mass for Silt and assumption of a soil mass failure for Sand ). It is noted that the analytical methods provide a good approximation with Tmax. The numerical applications were carried out with values of K0 derived from the simulations (K0=0.6 in Silt, K0=0.7 in Sand).

In most design methods in the literature, Tmax is considered to be proportional to tan( g) and it is worth reminding that ( g) is the same for the two soils . Both experimental and analytical results, however, show a differential behaviour. On the other hand , values for T max obtained by traditional methods ( 4.2 ) for the L-shapeanchorage are also calculated: Method (A) : T max = 30.4 kN/m (silt) T max = 34.5 kN/m (sand) Method (B) : T max =107.85 kN/m (silt) T max = 105.6 kN/m (sand)

That is not at all surprising that maximal pull-out forces are poorly dependent of the soil type, since interface friction is of the same order for the two different soils. The method (B) overestimated significantly the experimental values. This could be partially attributed to the combination of two assumptions: firstly neglecting the reduction of the vertical force acting on the last horizontal segment (3-4) (Figure 39) far less large than the weight P2 ,secondly taking the bend effect (not taken into account by Koerner) which magnified the pull-out force ( weighting coefficient K ) at every bend . Figure 42. Experimental pull-out forces for a L-shapeanchorage embedded in two different soils, and comparison with design methods (A) and (B) and proposed analytical method.

0

10

20

30

40

50

0,00 0,05 0,10 0,15 0,20U 0 (m)

T1 (k

N/m

)

Analytical - =34Experimental - =34

Silt

method A : T1 = 30,4 kN/mmethod B : T1 = 107,8 kN/m

0

5

10

15

20

25

0,00 0,05 0,10 0,15U 0 (m)

T1 (k

N/m

)

Analytical - =34Experiment - =34

Sand

method A : T1 = 34,5 kN/mmethod B : T1 = 105,6 kN/m

144

5 USE OF INCLINED PLANE FOR EROSION CONTROL EXPERIMENTATIONS Figure 43. Assessment of rain erosion control by geosynthetics

The use of geosynthetics for rain erosion control of civil engineering earth works is increasing (Reiffsteck, 2003). As a matter of fact, geosynthetics manufacturers are likely to propose a very large range of materials, but previously the civil engineer has to analyse the local conditions of erosion process and to select the best adapted geosynthetic structure. At least two problems have to be considered (Faure et al, 1996 ) : The first problem is to stabilise a significative thickness (several centimeters) of generally vegetative (or

sometimes cohesionless) soil on a steep slope. This problem will be call meso-stability in contrast with the macro-stability or overall stability of the earthwork.

The second problem is related to the grain and seed stability on the surface of the protection layer mentioned above.This problem will be call micro-stability. When the meso-stability is linked to static gravity forces, the micro-stability is related to the well-known erosion solicitations, rainfall splash effect and runoff. And generally, we need : - permanent solutions for meso-stability - temporary solutions (before vegetalisation) for micro-stability.

To participate to erosion control, geosynthetics manufacturers propose many products which can be classified in three families depending on the functions assumed by the product. 5.1 Classification of geosynthetics for erosion control:

Geocellular Confinement Systems (GCS):

These geocells are cellular confinement systems. The size of this honeycombs structure can be modified (height and width of the cells). The function is essentially the meso-stability. The geocells prevent any mass movement of the topsoil layer by providing tensile reinforcement. Anchoring and possibly nailing by pegs (excepted in case of lining system) increase the sliding stability. On the other hand the geocells, with large openings face to the rain do not protect against splash effect and the limitation of the runoff will be correlated to the width of the geocells.

Turf reinforcement mats (TRM):

These geomats have a three-dimensional structure, by association of several grids or random arrangement of threads (many different structures are available). The thickness is around one ore two centimeters much less than for geocells, the faces are flats and the apertures are large enough to be easily filled by soil and seeds. As for geocells, the main target is to stabilise an overlaid topsoil to vegetalize. However it is necessary to take care of the tensile strength level of these geomats, compatible with the tangential sliding force to equilibrate. If

Erosion Control

145

meso-stability is checked, TRM can play a complementary role in the micro-stability in contrast with geocells : the network of fibers can partially protect against splash-effect and act as a roots net against runoff. However the quality of the interface between geomat and subsoil (to avoid preferential flow) is a big concern, and a high flexibility of the mat is required is the subsoil support is not regular Erosion protection fabric (EPF) :

These materials are often called "Erosion Control Revegetation Mats" (ECRM) but in our opinion, this term is no enough accurate. The function of the mat is more simple. It is only a superficial protection against rainfall. The EPF is placed above the seeded subsoil. Its apertures can be smaller than for the TRM above, but must allow the growth of the plants through the mat. Only low tensile strength is required since meso-stability of an additive soil layer is not concerned. Variable 2D structures in polymeric or natural constituent are suitable for this application (grids, wovens, non-wovens,...). But in this case, still more, a close contact between the fabric and the subsoil is difficult to obtain on account of the lightness of the sheet, in the lack of soil included inside or above the EPF to make it heavier.

5.2 Specific Inclined Plane for erosion experimentations:

This facility was constructed at the Lirigm in Grenoble (Figure 44). It is composed of an inclined plane (2 meters long and 1 meter wide). Its slope angle can vary between 0 an 60. An element of soil of about 0.30 m thick can be tested.

An upstream water reservoir allows to observe runoff effects. This runoff water flow is controlled by a flowmeter upstream.

The rainfall simulator apparatus is made up with 22 pipes (1.30 m long and 0.10 m spaced) in which 264 watering droppers are pined every 0.10 m. The pipes are fixed on a metallic frame interdependent of eccentric wheels which turn with a variable velocity during the test. So whatever the droppers describe a 0.14 m diameter circle, the falling drops do not describe always the same circle due to the centrifugal force variations. The mean size of the drops is about 5 mm, but as the droppers are the same, the drop size distribution is uniform. Figure 44. Lirigm rainfall erosion simulator

Runoff

Rainfall simulator

water

Soil

Geosynthetic

pipes with droppers

146

5.3 Erosion control experimentations:

A serie of tests was done with an alluvial soil of the Isre river valley. It is a grey fine sandy silt (70% smaller than 80 m). Its uniformity coefficient (d60 / d10) is large, about 17.

A rainfall intensity of 100 mm/h was applied during 20 minutes.The splash effect is combined with a run-

off of variable intensity.The erosion is characterized by the mass of eroded soil after 20 mn of rainfall.This type of facility is versatile enough to simulate the principal critical situations for erosion.

The structures tested in the present case are the bare soil without vegetation as reference, a biodegradable

woven geojute, two geomats which are 3D materials with one (T1) or two (T2) plane grids as support and a shower wavy grid.The sheets are not filled with soil.

The Figure 45 displays the evolution of the erosion (soil mass transported) versus the value of the upper

run-off, for a constant slope = 15 and a constant rainfall.A typical erosion mechanism is demonstrated on this figure: the erosion due to splash effect is counteracted by the thickness of the run-off water.On the other hand all the sheets are efficient to reduce the erosion. However the wet geojute loaded by absorbed water is in these circumstances more efficient than geosynthetics, due to the better conformity of geojute with soil support. Below the jute ribbons, the soil is well protected. Into the meshes, micro terraces occurred where water is retained. This water produces an important protection to the rain impact and slows down the water flow.

The Figure 46 shows that soil erosion does not increase very much between = 15 and = 30.On the

other hand , for =45, there is a significative increase of the erosion for every product. For T1 , which has not a good contact with soil support, erosion is greater than for bare soil.This demonstrates that high flexibility combined with sufficient mass per unit area are required to obtain good protection. Figure 45 . Inclined Plane facility: Soil erosion for a slope = 15 , increasing run-off and different erosion

control products.

0

1000

2000

3000

4000

5000

6000

0 220 440 660 880

Runoff (l/h)

Eros

ion

(g/m

)

Bare soil at 15T1 at 15T2 at 15Jute at 15

147

Figure 46. Soil erosion for different slopes, different erosion control technics a run-off of 220 l/h and a rainfall of 100mm/h.

6 CONCLUSIONS

Stability of Geosynthetic Lining Systems is , for a geotechnical standpoint, a complex matter. Three geomechanical questions were identified: sliding of the geosynthetic lining system on slope, pull-out strength of the geosynthetic anchorage at the top of the slope , rain erosion of the cap cover .

Some research programmes on these topics, carried out in France and more specifically in Lirigm-Grenoble University, are presented. They include laboratory tests and large scale experimentations on actual slopes .An extensive use of Inclined Plane facilities with different configurations demonstrates the potential profit of this kind of device.

The observations derived from the tests and their detailed interpretation are really fruitful ,as they highlight specific local interaction behaviour between soil and geosynthetics, which are not taken into account in design methods , more particularly wrinkles of the geosynthetics and complex distribution of the real relative displacements of the components of the Geosynthetic lining system along the slope , realistic value of interface friction angle , pull-out strength of a L-shape anchorage ,identification of the mechanisms of control of erosion by geosynthetics from these results, some proposals for alteration of the current design methods are made.

It s worth noting that, although focused on landfill slope barriers, in most cases the results presented are also easy to apply for example to road embankments slopes and river banks.

7 REFERENCES

Brianon, L., Girard, H., Poulain, D., and Mazeau, N., Design of anchoring at the top of slopes for geomembrane lining systems , 2nd European geosynthetics conference, Bologna, Italy, 15-18 October 2000, 2: 645-650 (2000)

Brianon, L., Girard, H. and Poulain D., 2002, "Slope Stability of Lining Systems Experimental Modeling of Friction at Geosynthetic interfaces", Geotextiles and Geomembranes, Vol. 20, pp. 147-172

Chareyre, B., Brianon, L., and Villard, P., Theoretical versus experimental modelling of the anchorage

capacity of geotextiles in trenches , Geosynthetics International, 9(2): 97-123, (2000)

Chareyre, B., and Villard, P., Dynamic Spar Elements and DEM in 2D for the modelling of soil-inclusion problems Journal of Engineering Mechanics ASCE (2004)

01000200030004000500060007000

0 15 30 45 60

Pente ()

Eros

ion

(g/m

) Bare soilT1T 2Jute

148

Faure,Y.H.,Gourc,J.P.,Ratel,A., Use of geosynthetics in slope erosion control, 1st European Conference on Erosion Control, Barcelona , Spain , (1996)

Feki,N.,Villard,P.,Gourc,J.P.,Chatelet,L.,Gisbert,Th., Comparative long term survey of geosynthetic cap

lining systems, Proceedings of the 7th International Conference on Geosynthetics, Nice , France ,Vol 2 pp 663-666 , (2002)

Giroud J.P., Beech J.F., Stability of soil layers on geosynthetic lining system, Proceedings of Geosynthetic89

Conference , San Diego, USA, (1989) Gourc ,J.P.,Lalarakotoson,S.,Muller-Rochholz,H.,Bronstein,Z., Friction measurement by direct shearing or

tilting process-development of an European Standard Proceedings of Eurogeo 1,1st European Conference on Geosynthetics, pp.1039-1046 , Maastricht , Nl (1996)

Gourc J.P., Berroir G., Stock R., Begassa P., Assessment of lining systems behaviour on slopes,

Proceedings of Sardinia 1997, Sixth International Landfill Symposium , Cagliari, Italy, 11 p., (1997) Gourc,J.P.,Villard,P.,Reyes-Ramirez,R.Feki,N.,Brianon,L.Girard,H. Use of inclined test to assess stress

mobilization of liner on slope ,Proceedings of Kyushu 2001,Symposium on Reinforcement of soils , pp.201-205, Kyushu,Japan (2001)

Gourc,J.P.,Reyes Ramirez, R., Dynamic-based interpretation of the interface friction test at the inclined

plane, Geosynthetics International (to be published , Vol.11) (2004) Guide technique, 2000, Etanchit par gomembranes des ouvrages pour les eaux de ruissellement routier,

co-dit par le SETRA et le LCPC, guide complmentaire, 71p , (in French) , (2000) Imaizumi, S., Tsuboi, M., Doi, Y., Shimizu, T. and Miyaji, H., Anchorage ability of a geosynthetics liner

buried in a trench filled with concrete, Sixth International Landfill Symposium, Sardinia 97, Cagliari : Environnemental Sanitary Engineering Centre, Italia, 13-17 october 1997, pp. 453-462 , (1997)

Koerner, R.M., Designing with geosynthetics Prentice Hall, Upper Seddle River, New Jersey, USA, Fourth

Edition: 487-494 , (1998)

Koerner R., Hwu B.L. Stability and tension considerations regarding cover soils on geomembrane lined slopes Geotextiles and Geomembranes , Vol 10, pp. 335-355, (1991)

Lalarakotoson,S.,Villard,P;,Gourc,J.P., Geosynthetic Lining systems reinforced by geosynthetic inclusion

Proceedings of the 6th International Conference on Geosynthetics , Atlanta,USA , pp.487-490 (1998) Lala Rakotoson, S.J., Villard, P. and Gourc, J.P., "Shear Strength Characterization of Geosynthetic Interfaces

on Inclined Planes", Geotechnical Testing Journal, Vol. 22, No. 4, pp. 284-291, (1999) Olivier,F.,Gourc,J.P.,Lopez,S.,Benhamida,S.,Van Wyck.D., Mechanical behaviour of solid waste in a fully

instrumented prototype compression box, Proceedings Sardinia 2003, 9th International Landfill Symposium, Cagliari , Italy (2003)

Palmeira, E.,Viana,H., Effectiveness of geogrids as inclusions in cover soils of soils of slopes of waste

disposal areas, Geotextiles and Geomembranes,Vol.21, pp 317-337, (2003) Poulain,D.,Girard,H.,Glaud,V.,Haddane,K.,Stability and anchorage of geosynthetic systems on

slopes :development of a new designing tool, Proceedings of the 3rd European Conference on Geosynthetics,Eurogeo 3 , (2004)

Reiffsteck,Ph.,Geosynthetiques et controle de lrosion (in French)- 13th Regional Conference African on

Geotechnics, Marrakech, Marocco , (2003)

149

Reyes-Ramirez,R.,Thomas,S.,Feki,N., Stability of different inclined cap liner systems-Landfill field trial ,

Proceedings of Eurogeo 2, 2d European Conference on Geosynthetics, Bologna, Italy,Vol.2,pp. 523-526 (2000)

Reyes Ramirez,R.,Gourc,J.P.,Billet,P., Influence of the friction test conditions on the characterization of the

geosynthetics interfaces ,Proceeding of the 7th International Conference on Geosynthetics, Nice , France, (2002)

Reyes-Ramrez, R. and Gourc, J.P., "Questions Raised Regarding Use of the Inclined Plane in Measuring

Geosynthetic Interface Friction Relationship", Geosynthetics International , Vol.10, (2003) Soyez B., Delmas Ph., Herr Ch., Berche J.C., Computer evaluation of the stability of composite liners,

Proceedings of the 4th International Conference on Geosynthetics , Balkema, Rotterdam, pp. 517-522 , (1990)

Standard prEN ISO 12957-1, 2001-2002, Geotextile and geotextile-related products Determination of

friction characteristics, Part 1 : Direct shear test, European Committee for standardisation (CEN), European Standard, Brussels, Belgium (2001-2002)

Standard prEN ISO 12957-2, 2000, Geosynthetic Determination of friction characteristics, Part 2 : Inclined

plane test, European Committee for standardisation (CEN), European Standard, September 2000, Brussels, Belgium, 11p (2000)

Villard P., Gourc J.P., Feki N.. Anchorage strength and slope stability of a landfill liner, Proceedings of the

Conference Geosynthetics97, Vol. 1, Long Beach, California, USA, March 1997, pp. 453-466 , (1997) Villard,P.,Gourc,J.P.,Feki,N.,Analysis of geosynthetic lining systems undergoing large deformations,

Geotextiles and Geomembranes, Vol 17,pp 17-32, (1999) Villard,P. Chareyre,B., Design methods for geosynthetic anchor trenches on the basis of true scale

experiments and Discrete Element Modelling , Canadian Geotechnical Journal, (2004)

ContentsPrintRetune